一种基于最小1-范数准则的TDOA估计算法

An Estimation Algorithm of TDOA Based on Least 1-Norm Criterion

针对脉冲噪声条件下利用传统广义互相关法(Generalized Cross-Correlation,GCC)进行时延(TDOA,Time Difference of Arrival)估计性能退化问题,提出一种基于最小1-范数准则的TDOA参数估计算法。对于高斯噪声,传统GCC估计方法能够实现统计最优,但当噪声的统计分布为非高斯分布时,利用传统GCC参数估计方法的估计精度和鲁棒性急剧下降。利用最小1-范数准则,提出一种存在α-稳定分布重尾脉冲噪声环境下的TDOA估计算法。系统仿真实验与结果分析表明,与传统GCC方法和分数低阶矩(Fractional Lower Order Moments,FLOM)方法相比,该算法在鲁棒性和估计精度方面均有明显改善。

To solve the problem of the performance degradation of TDOA （ Time Difference of Ar- rival） estimation using generalized cross-correlation （GCC） method under impulsive noise con- ditions,an estimation algorithm of TDOA based on least 1-norm criterion is proposed. For the Gaussian noise, the traditional GCC estimation method can achieve statistical optimization, but when the statistical distribution of noise is non-Gaussian distribution, the performance of the tra- ditional GCC method decreases drastically in the estimation accuracy and robustness. The con- cept of 1-correlation function is given by using the minimum 1-norm criterion, and a TDOA esti- mation algorithm with oL-stable distribution noise of heavy-tailed distribution is proposed. Simu- lation results of the system model show that the performance of the proposed algorithm is better than that of the traditional GCC method and fractional lower order moments （FLOM） method in terms of robustness and estimation accuracy.